Semantic Search
Last updated
Last updated
The IGOR^AI semantic search employs advanced semantic search algorithms that analyze the meaning and context of user queries to retrieve relevant scientific abstracts. It utilizes techniques such as embeddings, semantic similarity measures, and context-aware representations to identify abstracts that closely match the user's information needs.
To run an IGOR^AI semantic search, you can write your text as a question. IGOR^AI will then retrieve a list of papers that are sorted by relevancy to your inputs.
The different parameters from the results are:
Link: allows you to get to the source document
Relevancy: Informs you on the accuracy of the document related to the query. The higher the relevancy, the more likely your document answers or is connected to your questions.
Title: Provides you the title of the document
Similars: By clicking on similar, you will be able to find more documents that are similar to the selected one
Abstract: The original text of the abstract
Date: Publication date of the document
The system employs disambiguation techniques to address ambiguity or polysemy in user queries by considering the context in which terms are used and leveraging knowledge about the specific domain or research area. This helps ensure that the retrieved abstracts are relevant to the intended meaning of the query.
The relevance of retrieved scientific abstracts is determined based on various factors, including semantic similarity to the user query, the quality and recency of the abstracts, relevance scores computed through machine learning algorithms.
Our system is designed to handle noisy or incomplete queries. This helps improve the accuracy and relevance of the retrieved abstracts, even when the query is ambiguous or lacks specificity.
